EP3687627A2

EP3687627A2 - Method for epid-based verification, correction and minimisation of the isocentre of a radiotherapy device

Info

Publication number: EP3687627A2
Application number: EP18814769.8A
Authority: EP
Inventors: Markus Wösle
Original assignee: Staedtisches Klinikum Dessau
Current assignee: Staedtisches Klinikum Dessau
Priority date: 2017-09-27
Filing date: 2018-09-27
Publication date: 2020-08-05
Anticipated expiration: 2038-09-27
Also published as: PT3687627T; US11607564B2; PL3687627T3; DE102017009040A1; DE112018005512A5; CN111432878B; WO2019063035A3; HUE057803T2; CN111432878A; ES2909138T3; US20200289850A1; EP3687627B1; WO2019063035A2

Abstract

The invention relates to a method for EPID-based verification, correction and minimisation of the isocentre of a radiotherapy device, in which method, according to the following steps: a) a measurement body is positioned, by means of the projection device, at the projection position in the current radiological isocentre of the radiotherapy device, b) an irradiation field, limited by the collimator, is applied for at least one pre-defined angular adjustment of the support arm, the patient couch and the collimator and c) at least one common dose image of the measurement body and the irradiation field is captured using the EPID, d) a dose profile for each direction within an EPID co-ordinate system is created on the basis of the captured dose image and e) in the plot of the dose profile an inflection point between a local dose minimum and a local dose maximum, and between a local dose maximum and a local dose minimum is determined at each of both expected bodily limits of the measuring body in the X-direction of the EPID-co-ordinate system and at each of the two expected bodily limits of the measuring body in the Y-direction of the EPID-co-ordinate system, and f) the determined positions of the inflection points are linked to the bodily limits of the measurement body in the X-direction and in the Y-direction, g) the position of the centre point of the measurement body relative to the EPID-centre is determined in the dose image on the basis of the linked bodily limits of the measurement body, the steps d) to g) being carried out in the same way for the field limits and the field centre point of the irradiation field, and h) a differential vector is determined from a deviation in position of the centre point of the measurement body from the EPID-centre and from a deviation in position of the field centre point of the irradiation field from the EPID-centre, and i) the vector components of the differential vector are used to correct the current radiological isocentre.

Description

Procedures for EPID-based verification, correction and minimization of the

Isocenter of a radiotherapy device

The invention relates to a method for checking, correcting and minimizing the isocentre of a radiotherapy device, which comprises at least one patient support table rotatable about at least one table axis, a support arm rotatable about a support arm axis, a radiation head arranged on the support arm for generating a therapy beam, a rotatable collimator, a device for the projection of the radiological isocenter and a portable arm-mounted digital recording system (EPID, English: Electronic Portal Imaging

Device) for detecting dose images by means of the therapy beam comprises.

In the treatment of tumor diseases by means of ionizing radiation, an essential challenge is to administer the therapeutic energy dose to the focus of the disease and to protect the neighboring healthy tissue in the best possible way. This is achieved by mechanically and theoretically precise radiotherapy devices, such as electron linear accelerators, gamma knifes, cyberknifes, proton and heavy ion accelerator systems in conjunction with safety margins adapted to machine and patient positioning tolerances in the target volume definition. A central feature of any radiotherapy device is the spatial deviation of the central ray from the ideal punctate site, called the isocenter or ideal isocenter. In an ideal radiotherapy device, in which no mechanical tolerances occur, all angle-dependent central rays intersect in this isocenter. medical

Electron linear accelerators, which are used in almost every radiotherapy institution for tumor therapy, have at least three rotational degrees of freedom. These include the support arm angle with a usual value range of -180 ° to + 180 °,

the collimator angle with a usual value range of -175 ° to + 175 ° and

the isocentric angle of the patient table with a usual value range of -95 ° to + 95 ° For a real radiotherapy device, the isocenter is not a point, but an isocentroid with spatial extent (hereinafter also referred to as spatial isocenter). The term centroid is derived from ellipsoid, which is a body formed by rotation of an ellipse about one of its axes. The central rays of all combinations of the above angles intersect or touch the isocentroid. To determine the global

Isozentroids a radiotherapy device are initially the respective size and location of the individual isocentroids depending on each individual degree of freedom to determine. This measurement and verification of the isocentroids is also known as Winston-Lutz test or Winston-Lutz method. In the Winston-Lutz test, a radiopaque or radiation-tight measuring body, which is also called the Winston-Lutz pointer and is rigidly coupled to the patient table of a radiotherapy device, is first positioned in the radiological isocenter of a radiotherapy device. This isocenter is displayed with the aid of spatially fixed, visible line lasers or projected onto the Winston-Lutz pointer. Following the positioning of the measuring body, in each case a flat dose image is formed with a square or round irradiation field which is limited by the block apertures of a collimator, the lamellae of a multi-leaf collimator (MLC) or by a round collimator additionally mounted on the radiator head with various support arm angles, collimator angles and table angles using a digital recording system (EPID). Subsequently, a digital image processing of the prepared dose images is carried out, wherein different algorithms are used to determine the spatial position of the central beam relative to the measuring body can. The known methods and algorithms are based, for example, on digital high-pass filters, two-dimensional operators for edge enhancement and edge finding (such as Sobel operators, Canny filters), line extraction (Hough transformation), center of mass calculation (Signal Intensity Weighting). , the convolution method, the object segmentation (segmentation, Contouring) as well as on the thresholding method. In the review "Isocenter verification for linac-based stereotactic radiation therapy: review of principles and techniques" Rowshanfarzad, P _.; Sabet, M., O'Connor, DJ, Greer, PB (Journal of Applied Clinical Medical Physics, Volume 12, No. 4, 185-195, 2011) discusses the above methods and discusses their pros and cons. The major disadvantages of the EPID-based methods for isocenter verification are compared with film-based methods in a relatively lower resolution, making accurate verification of the radiological isocenter is difficult. In a recent publication, "Ravindran, PB (Australasian Physical and Engineering Sciencifics in Mediane, Volume 39, Issue 3, 677-685, 2016)," A study of Winston-Lutz test on two different electronic portal imaging devices and with low energy imaging " ), the Winston-Lutz pointer is moved real with the minimum step size of 0.25 mm, and from this it is determined by morphological operations a maximum error of the central ray distances of the Winston-Lutz pointer of 0.04 ± 0.02 mm It is known by means of Hough. Du, W., Yang, J. (Phys. Med. Biol., 54 (3), 555-567, 2009) Transformation in computationally varied dose images to achieve a theoretical accuracy of 0.02 ± 0.01 mm, with machine tolerances are not taken into account.

Other problems of the known methods consist in the impossibility of performing a measurement at certain Tragarmwinkel- / table angle combinations and in an unfavorable signal / noise ratio.

In EPID-based methods, there is still the problem that evaluation methods, such as the Hough transformation, are unsuitable for MLC-shaped radiation fields, since the lamellae can not form even field boundaries with their ends due to finite positioning accuracy and transmission between the lamellae.

Another disadvantage is that the positioning inaccuracies of the recording system in all EPID-based methods - in particular a violation of orthogonality to the central beam and a not exact distance to the isocenter - affect the outcome. Other deficiencies in the known EPID-based methods include a lack of representation of the global isocentroid, in which the central isomorphic deviation from the ideal isocenter for any combination of the three angles for the arm, collimator, and patient table is a lack of mathematical description of the global isocentroide to minimize it by optimization of parameters of the radiation therapy device as well as missing investigations of the influence of all measuring conditions on the result.

The object underlying the present invention is now to propose an EPID-based method for checking and correcting the isocenter of a radiotherapy device, with which the disadvantages known from the prior art can be avoided or reduced.

According to the invention the object is achieved by a method having the features of claim 1. Advantageous embodiments and further developments of the invention can be realized with features described in the subordinate claims. According to the object of the invention, a method for EPID-based checking, correction and minimization of the isocenter of a radiotherapy device comprising at least one about a table axis rotatable patient table, a pivotable about a Tragarmachse support arm, arranged on the support arm radiator head for generating a therapy beam, a rotatable Collimator, comprising means for projecting the radiological isocenter and a digital acquisition system (EPID) for acquiring dose images by means of the therapy beam, proposed as follows:

a) A measuring body is positioned by means of the projection device at the projection position in the current radiological isocenter of the radiotherapy device,

b) an irradiation field bounded by the collimator is applied for at least one predetermined angular adjustment of the support arm, the patient support table and the collimator, and thereby

c) with the EPID at least one common dose image of the measuring body and taken of the radiation field,

d) based on the recorded dose image, a dose profile for each direction is created within each EPID EPID coordinate system; and e) in the course of the dose profile, at both expected body boundaries of the measurement body in the X direction of the EIPD coordinate system and at both expected body boundaries of the measuring body in the Y-direction of the EPID coordinate system between a local dose minimum and a local maximum dose and between a local maximum dose and a local dose of minismium each determined a turning point and

f) the determined positions of the inflection points are assigned to the body boundaries of the measuring body in the X direction and in the Y direction,

g) based on the associated body boundaries of the measuring body, a position of the center of the measuring body relative to the EPID center is determined in the dose image, wherein the steps d) to g) are carried out analogously for the field boundaries and the field center of the irradiation field, that is, it is Based on the recorded dose image, a dose profile is created and in the course of the dose profile is at both expected field boundaries of the irradiation field in the X direction of the EIPD coordinate system and at both expected field boundaries of the irradiation field in the Y direction of the EPID coordinate system between a local dose minimum and a local maximum and between a local Dosismaximum and a local Dosisminium each determined a turning point and the determined positions of the inflection points of the field boundaries of the irradiation field in the X direction and assigned in the Y direction and based on the associated field boundaries of the irradiation field in the dose image one layer of the field center of the irradiation field relative to the EPID center,

h) a difference vector is determined from a positional deviation of the center of the measuring body from the EPID center and from a positional deviation of the center of the irradiation field from the EPID center, and i) the vector components of the difference vector are used to correct the current radiological isocenter.

In step d), a unit change from pixel to mm can be made to determine absolute distances. The unit change represents a transition from the digital to the analog dose profile, so that a terverarbeitung the geometry data can be ensured in an algorithm.

The coordinate system which is essential for the measurement setup and for carrying out the method according to the invention is the inertial system (space-fixed radiotherapy device coordinate system), in which the isocenter deviations (central beam deviations from the ideal isocenter) are indicated. The inertial system has the ideal isocenter ISO (center of the measuring body or the tungsten sphere of the Winston-Lutz pointer) as the origin. Since the coordinate system of the recording device used for steps c) to h) and the collimator coordinate system are movable, coordinate transformations are required. In the variation of the Tragarmwinkels (G) of the respective difference vector is imaged by means of the left matrix in the equation (5) below in the Intertialsystem, where ß = G is. In the variation of the collimator angle (C), before the analysis, the dose images are mapped into the EPID coordinate system using the right-hand matrix in equation (5), where y = -C; after analysis, the vectors with the same matrix, where y = C, are mapped into the inertial system. When varying the patient table angle (T) with G * 0 °, the left matrix in equation (5) with ß = G for coordinate transformation is necessary again. Since errors occur as a result of machine tolerances - in particular of positioning errors of the recording system - all differential vectors are imaged from the real to the ideal coordinate system before being imaged into the inertial system according to equation (2) below. Also of relevance is the EPID coordinate system, which is positioned in a support arm-proof manner and has its origin in the central detector element of the EPID. Preferably, the projection of the current radiological isocenter is carried out by means of five space-stable line lasers. With crosshairs, which may be executed in the form of color-highlighted notches on the surface of the measuring body, the measuring body in the projected isocenter of the radiotherapy device can be accurately positioned.

According to an advantageous development of the method according to the invention, the method steps b) to h) for the angular degrees of freedom of the patient support table, the support arm and the collimator can be provided with an increment of not more than 30 °. Expediently, it is also possible to select those angle settings which are relevant for a subsequent treatment of a patient. It may further be provided that from dose images of different

Angular positions of the patient table, the support arm and the Collimators originate, differential vectors can be used to determine the size and position of the spatial isocenters are used, the vector components of the position vectors of the spatial isocenters are used to correct the radiological isocenter.

The method according to the invention can preferably be carried out automatically, using appropriate software. An automatic performance of the steps d) to i) and in particular an evaluation processing of dosage profiles for finding inflection points as well as performing coordinate transformations can be carried out by means of routines, for example in the software package MATLAB ^®.

Preferably, in step e), the dose profile of the dose image in the X direction and in the Y direction can be examined for inflection points between a dose minimum and a maximum dose and in the further course between a maximum dose and a minimum dose to all body boundaries of the measuring body and limits To determine the radiation field. Ideally, the measuring body is a tungsten sphere with a diameter between 5 mm and 10 mm. It should be noted that those found in the failure mode analysis and error influence analysis

Parameter values for suitable measurement conditions and analysis conditions are preferably valid for a ball diameter of 5 mm. However, the method according to the invention should not be limited to a spherical measuring body in order to be carried out. Therefore, other body shapes of the measuring body are also conceivable.

According to one conception of the invention, the two-dimensional mapping of the diameter of a spherical measuring body and the field width of an irradiation field in the dose image corresponds to the distance between two inflection points in the dose profile parallel to the X direction of the collimator fixed diaphragm coordinate system. The points of inflection can be those points which have exactly 50% of the absorbed dose of the field center in the dose image. Accordingly, in the Y direction, those points can be regarded as turning points, which have exactly 50% of the absorbed dose of the field center in the dose image. The inflection points determined are assigned to the field boundaries of the irradiation field. Since the local dose minimum is located behind the measuring body in the area of the two-dimensional field center, the smallest of the local dose maxima of the dose profiles in the X direction and Y direction can be used as a replacement for the 100% dose. The determination of the body boundaries of the measuring body is analogous. The absorbed dose to define the boundaries of the spherical measuring body is the arithmetic mean of the 100% replacement dose and the local dose minimum behind the ball, ie a 50% replacement dose. The dose maxima are located between the spherical measuring body and the field edges, which are defined by the collimator setting. According to the above statements, it can be provided that the inflection point (s) is / are determined in the range of a 50% dose point between a dose minimum and a maximum dose level and / or between a maximum dose level and a dose range of the dose profile. The definition of the field size and the central beam position with the help of the 50% -sodose is described in the standard IEC 60976.

The measurement planes perpendicular to the image plane, which cut out the corresponding dose profiles for defining the field size, the central ray and the sphere size and the sphere center, can be automatically determined by means of software using an algorithm. Starting from the local dose minimum in the dose image of the spherical measuring body, which lies approximately in the image center, two orthogonal measurement planes can be defined in a first iteration step. The determined dose profiles are analyzed with respect to field size, central ray position, sphere size and center of the sphere. In the second iteration step, the symmetry lines of the irradiation field and of the spherical measuring body can be determined with the ascertained central beam position and the determined ball center of the measuring body. In the case of irradiation fields limited by means of a circular collimator and in the determination of the spherical center of the measuring body, the lines of symmetry are identical to the lines of intersection of the measuring planes. The above-described analysis of dose profiles or the assignment of inflection points to body boundaries and field boundaries provides the distances of the central jet piercing point and the ball center of the measuring body in the image plane relative to the EPID center.

It has been shown that with the method according to the invention in the definition of the field size, the central beam position, the ball diameter and the center of the spherical measuring body, a spatial resolution of 0.01 mm for the spatial deviation of the central beam can be achieved by the spherical measuring body, which can be detected by measurement. For example, a process-related resolution of 0.01 mm could be achieved using an EPID with a pixel size of 0.392 mm. This can be achieved by considering a section of the inflection tangent of the dose profile in the region of the 50% dose point for the exact determination of the field boundaries or the body boundaries of the spherical measuring body. In this area of the dose profile, the turning tangent is the best approximation of the dose courses over the location coordinates. The least error approximation can be achieved with the shortest tangent still present in the discrete pixel size digital dose image. In the direction of the location coordinate, it measures exactly one pixel - or, as a special case, two pixels if the 50% dose point coincides exactly with one pixel center. The two pixels in the range of the 50% dose point are determined, one of which has a lower and one greater gray value than the 50% dose point. In this case, with knowledge of the pixel size, a specific distance can be determined. The exact location coordinates for the field boundaries of the radiation field and the body boundaries of the spherical measuring body can be achieved by linear interpolation between the two excellent pixels.

According to the above-described advantageous embodiment of the method according to the invention, it can be provided that the inflection point (s) in the range of the 50% dose point is / are preferably set between two pixels, of which a first pixel has a lower dose than 50% and one on the first pixel adjacent second pixel represents a larger dose than 50%.

Finally, the difference vector is calculated as a displacement vector between the central jet piercing point and the center of the spherical measuring body, measured in the EPID plane, by means of a centric elongation in FIG the isocenter plane is mapped, where the stretch center is the beam focus and the stretch factor is

depends on the constants "focus center distance" SAD and the variable "focus EPID distance" SID. In addition, the dependent on Tragarmwinkel

Position vectors imaged by means of a coordinate transformation from the image plane into the spatially fixed coordinate system of the radiotherapy device. The orientation of this inertial system is specified in the IEC 61217 standard. According to another concept of the invention, the spatial vectors of the spatial isocenters are used to calibrate a patient positioning system. Alternatively or additionally, provision may be made for the position vectors of the spatial isocenters to be used for the correction of the projection device in order to be able to adapt the laser projection of the radiological isocenter. The location vector of the isocentroids is to be understood as the spatial center position of the isocentroids.

The advantage of the method according to the invention lies in particular in the extraction of a dose profile with the unit of length mm from a dose image with the unit of length pixels, so that this with less computational effort in a shorter time and in particular with computationally higher

Resolution can be processed. Advantageously, it is utilized that a majority of the known radiation therapy devices anyway have a possibility for digital recording (EPID) of an irradiation field, so that an isocenter verification and a corresponding correction with existing devices can be carried out with little effort. So it is possible because of the relatively small amount of time to carry out and the automatic analysis of the method according to the invention that an Isozentrumsverifikation (review of the isocenter) can be performed immediately before a radiosurgical application to ensure the best possible patient safety. Furthermore, it has been found that a computational resolution of 0.01 mm can be achieved by the method according to the invention, without requiring any technical changes, such as high-resolution recording systems. In summary, with a To the invention carried out isocenter verification increased patient safety can be achieved with existing radiotherapy devices with EPID. Since comparatively less time is required to carry out the isocenter verification and isocenter correction according to the method according to the invention, costs can also be saved.

For further optimization, it may be provided according to a development of the method according to the invention that a correction of machine tolerances is made. For example, a correction of the position vectors in the EPID coordinate system can be provided. At a

Incorrect positioning of the EPID can be used to correct the position vectors further transformation step from the shifted EPID coordinate system to the ideal EPID coordinate system will be required:

where is the correction matrix

as a product composed of three elementary turns. According to the calculation rule "multiply vectors with matrices from the left", the rotation order is defined as for gimbals: angle α around the X-axis, angle β around the Y-axis and angle γ around the Z-axis In its acceptance test, all angular errors are ≤ 1 °, the rotation order has no influence on the correction, and it applies to a general angle with α in radians

Rotation around the X-axis of the ideally positioned EPID is the right-hand mapping matrix in Equation (3)

a function of the angular deviation a. Accordingly, in equation (3), the rotation matrices are about the Y-axis and Z-axis which depend on the corresponding angular deviations β and γ. The definition of the gimbal angles as well as the mapping matrices in equations (4) and (5) can be found, for example, in "Multibody dynamics with unilateral contacts" Pfeiffer, F., Glocker, Ch. (Wiley series in nonlinear science, series editors: Nayfeh, AH and Holden, AV, John Wiley & Sons, Inc., New York, 1996). Translational corrections do not have to be applied because the searched difference vector

between the puncture point of the central ray and the sphere center of the measuring body does not depend on displacements of the EPID in its XY plane. The tolerance of the vertical coordinate is taken into account in the centric extension from the image plane to the isocenter plane by the corrected focus EPID distance in equation (1). The tolerances of the three angular degrees of freedom can also be taken into account. Deviations measured for the bracket angle, the collimator angle, and the table angle can be taken into account when inputting appropriate software, as part of an algorithm for error correction.

In radiosurgery and stereotactic precision irradiation, due to irregular structures of the tissue to be irradiated, usually no rectangular or square irradiation fields are used. For such cases, the radiation fields are limited by means of the lamellae of a multi-leaf collimator (MLC). According to one embodiment variant of the method according to the invention, in which an MLC is used, therefore, the central beam positions between several pairs of lamellae of an MLC in the X direction are determined. Thus, it can be provided that, when using an MLC, steps d) to h) of the method according to the invention are carried out for each lamella pair of the MLC delimiting the irradiation field. It may further be provided that the outer pairs of lamellae are not used for the analysis, since there the dose were disturbed by scattered photons from the aperture system to the field boundary in the Y direction.

To determine the central beam position in the Y direction, two dose profiles are evaluated for irradiation fields which are formed by block apertures or the MLC, which are cut out in mirror image form to the symmetry line of the X coordinate. The mirror-symmetric measurement planes and the arithmetic averaging of the individual central beam positions in both directions X and Y have the advantage that any angular errors of the collimator are compensated for by the averaging. Another advantage is that fewer field disturbances are caused by the measuring body, since the interest of interest 50% dose points of the dose profiles have a greater distance from the measuring body or the Winston-Lutz pointer.

According to a further advantageous embodiment variant of the method according to the invention, a variation of the patient table angle occurs

Support arm angle 0 ° adjusted to obtain a statement about the vertical position change of the measuring body. Preferably, with a variation of the angle of the patient support table, a support arm angle of 30 ° is set, whereby an angle of the patient support table in the range between 0 ° and 90 ° is made possible. Similarly, a support arm angle of -30 ° allows an angle of the patient table in a range between 0 ° and - 90 °. In the event that a space-fixed, independent of the radiotherapy device patient positioning system is used, the Isozent- roid of the table can be determined alternatively without movement of the support arm. In the above variant, the table angle variation with the discrete

Angles (-90: 30: 90) °. The position of the support arm is 0 °, so that the radiator head is not in the two Nutzstrahlenfeldern the external X-ray system. In the event that the patient positioning system ExacTrac 6.0.6 (BRAINLAB AG, Feldkirchen, Germany) is used, the spatial position of the patient is set at each table angle

Measuring element, which may be a Winston-Lutz pointer from BRAINLAB AG (Feldkirchen, Germany), was detected radiologically in the verification step Detect Winston-Lutz Pointer with the functionality Winston-Lutz Pointer Analysis. The negative values of the displacements of the measuring body relative to the ideal isocenter of the radiotherapy device are vectorially stored in the file WL_Test.log. This is data in the form of a text document which can be read by the software package MATLAB ^® . In the data of the file goes from columns 15-17 of the table angle-dependent displacement vector

in the initial system the form

having. The offset

of the central beam from the ideal isocenter of the radiotherapy device is known from the determination of the support arm angle dependent isocentroide and does not change with the table angle. Thus, the displacement vector (distance vector) of the central beam from the measuring body is the vector sum

which describes the isocentroid of the patient table in the inertial system. According to a development of the method according to the invention, a global isocentroid can be determined from the individual deviations in the three spatial directions X, Y and Z. The procedure is as follows: After for each angular degree of freedom the geometry of the isocentroide with the characteristics

Deviations of the central ray from the Winston-Lutz pointer in the three spatial directions X, Y and Z as a function of an angle, Radial deviation of the central ray from the Winston-Lutz pointer as a function of an angle,

• location coordinates of the isocentroid,

• maximum diameter of the isocentroids in the three spatial directions and

• maximum global radius of the isocentroid

was determined in the spatially fixed coordinate system (inertial system), a global isocentroid can be determined with all results. This can be achieved by considering the angle-dependent measurement conditions in a summarized way:

• The isocentroid of the support arm is determined at 0 ° collimator and table angle.

• The isocentroid of the collimator is determined at 0 ° support arm and table angle.

Half of the isocentroid of the table is determined with 30 ° or -30 ° support arm angle and with 0 ° collimator angle.

Based on the isocentroid representation as a function of the support arm angle, the additional maximum possible isocenter deviations due to the collimator angle variation and table angle variation can be represented as scattering bands in the positive and negative directions above the support arm angle. The evaluation regarding the location coordinates and maximum diameter or the largest radius is finally updated. In addition, the center distances, which are to be measured at least every six months according to DIN 6875 - Part 2, are determined between all three axes of rotation. The scattering bands have a negative and positive component. Its determination first requires a mapping from the EPID coordinate system to the initial coordinate system for the isocenter deviations of the collimator angle variation:

The mapping matrix is known from equation (5), where β = G is the cantilever angle. The location vector is composed of the components of the isocenter deviation. The negative scattering bands arise too

Accordingly, one obtains the positive dispersion bands according to

Preferably, the interval of the collimator angle for the extreme search in equations (13) to (18) can be defined as C e [-90 °, + 90 °], since all optimal collimator angles in patient treatment are in this range. The scatter bands in the table angle variation are analogous to equations (13) to (18) by defining the collimator angle C through the table angle T and the extrema search intervals by T e [-90 °, + 90 °]. In the context of the process according to the invention isozentroids by

Optimization can be minimized. Minimization through optimization is conceivable for the

Isozentroid of the support arm in the event that a round collimator and an external patient positioning system is used,

Combined isozentroid of the support arm and the collimator in the event that a patient positioning system is used, • Combined isozentroid of the support arm and the patient table in the event that a round collimator without patient positioning system is used, and / or

• global isocentroid in the event that neither a round collimator nor a patient positioning system is available.

If an independent radiological patient positioning system is used in conjunction with the radiotherapy device, the isocentroid of the patient support table may go unnoticed because it detects and corrects all table angle dependent central beam deviations. When using circular collimators, the collimator angle C is usually C = 0 ° = constant.

When field limiting with an MLC, the optimization variables in the vector be summarized, where the centreline offset of the slats and y _{J0 is} the jaw offset of the Y- _pair of apertures. The first component of this vector may also be the jaw offset of the X-aperture pair, as long as it is intended to limit the irradiation field. The second component can also be identical to 0 if the field boundary in the Y direction is reached solely by the slats of the MLC. When using circular collimators, equation (19) includes their offsets. The isocentroids listed above, ie the isocentroid of the support arm, the combined isocentroid of the support arm and the collimator, the combined isocentroid of the support arm and the patient table, and the global isocentroid, are functions of the support arm angle G. The solitary support arm isocentroid is composed of the components

composed.

The combined Isozentroide can be described in any spatial direction by two features that they limit in the negative direction relationship ^¬, in the positive direction:

The collimator angle C and the table angle T of the patient support table can be varied within the specified interval. The value range of the collimator angle C can be limited to ± 90 °, since usually all optimal collimator angles are in this range.

According to a further advantageous embodiment variant of the method according to the invention, provision can be made for all device-specific parameters influencing the spatial isocenter to be optimized by using one or more predefined target function (s). Thus, a correction of machine parameters can be provided. Advantageously, the optimization can be carried out by means of software. Possible optimization functions that depend on the variables in equation (19) relate to the sums, arithmetic mean, extrema, and integrals over the squares of the squares of the squares

of the examined isozentroids at discrete Tragarmwinkeln

For the components of the spatial radius, the following applies:

at the discrete support arm angle G ,. The definition of diameter is

In the special case of the solitary isocentroid of the support arm, the geometric quantities that can be optimized are defined as follows:

with K e {X, Y, Z}. With the three geometric variables and four functions each, twelve different target functions can thus be selected for optimization, with some target functions being redundant with respect to the optimization:

In the search for the best optimization for a specific irradiation task, it is therefore not necessary to consider all objective functions, since the functions of each of the pairs mentioned here provide identical results.

During optimization, the position vector of the central ray deviations in equation (6) becomes varies in the detector coordinate system (EPID coordinate system). With the help of equations (5) and (6), the positional deviations of the investigated isozentroids in the spatially fixed coordinate system of the radiotherapy device can be determined. In equation (5), the angles β = G and γ = C are to be replaced by the support arm angle or the collimator angle.

The vector x ₀ is optimal if, for one of the objective functions outlined above, which contains the variables of equation (19) in quadratic form:

To adjust the collimator parameters on a radiotherapy device, the optimized parameters in equation (19) can then be selected from the various solutions. For example, those that have the mean spatial radius can be selected

minimize the central beam. For the systematics of identical, identical security rooms in all spatial directions, with which a tumor volume is extended to the irradiation volume, so that a tumor can also be completely irradiated with storage inaccuracies, tumor mobilities and machine tolerances, only the maximum diameter of the relevant isocentroids would be minimized:

In equation for solving the non-linear optimization problem without constraints (25) may be the subroutine fminsearch used in MATLAB ^®. The calculation scheme used here is a simplex algorithm of the Nelder-Mead type for a direct minimum search.

Furthermore, it may be provided according to an advantageous embodiment of the method according to the invention that geometric tolerances of the radiotherapy device are quantified and taken into account mathematically.

According to a further advantageous embodiment variant of the method according to the invention, provision may be made for the irradiation field to be under taking into account appropriate values of a field size, a relaxation time of the support arm, an energy-dependent dose per irradiation field, and / or a focus EPID distance is applied. Furthermore, an application of the method according to the invention to a therapy simulator can be provided, the spatial isocenters being determined, corrected and minimized.

The invention will be explained in more detail by way of example with reference to the following figures.

Show it:

FIG. 1 shows a two-dimensional dose image of a spherical measuring body in an MLC-limited irradiation field,

FIG. 2 a shows a dose profile of a two-dimensional dose image for explaining an embodiment of the method according to the invention,

FIG. 2b shows the dose profile from FIG. 2a for explaining the definition of FIG

50% -Dosispunkte,

FIG. 3 is a schematic representation of the metrological proof of the achieved spatial resolution, FIG.

FIG. 4 dose profiles of the Winston-Lutz pointer as measuring body,

FIG. 5 shows further dose profiles of an MLC-limited radiation field in the X direction,

FIG. 6 shows further dose profiles of an MLC-limited radiation field in the Y direction,

FIG. 7 shows a schematic representation of the geometry of a isotope-dependent isocentroid of a radiation therapy device, FIG. 8 shows a graphic representation of a global isocentroid.

To explain the method according to the invention, Figures 1, 2a and 2b are to be considered together.

In the inventive method for EPID-based verification and correction of the isocentre of a radiotherapy device, which at least a patient support table rotatable about at least one table axis, a support arm rotatable about a support arm axis, a radiation head for application of the therapy beam, a collimator for limiting a radiation field, a device for projecting the radiological isocenter and a digital acquisition system (EPID) for the generation of The method comprises the following steps: a) a measuring body 13 is positioned at the projection position of the radiological isocenter of a radiotherapy device,

b) then an irradiation field for at least one predetermined

Angle adjustment of the support arm, the patient table and the Collimator applied and thereby

c) recorded with the EPID a dose image of the measuring body 13, as shown in Figure 1. In the dose picture in Figure 1 is a two-dimensional gray-scale image of a dose MLC limited irradiation field size of 15 x 15 mm ² in MATLAB ^® (version R2007a). The EPID used has pixels with an edge length of 0.392 mm. The reference numeral 13 denotes a spherical measuring body (Winston-Lutz pointer) with a known diameter. The reference numeral 14 denotes a boundary of the irradiation field caused by the collimator. In the dose image, dark areas have a lower applied absorbed dose than bright areas. To determine the spherical center of the measuring body 13 by means of the crossed, orthogonal measuring planes 15, first both ball boundaries 1 and 2 in the X direction and both ball boundaries 16 and 17 in the Y direction are determined. The ball boundaries 1, 2, 16 and 17 can be determined from the dose profiles which are created on the basis of the recorded dose image in step d) of the method. FIG. 2 a and FIG. 2 b show a dose profile 5 in the X-direction of the dose image shown in FIG. The curve denoted by reference numeral 6 in FIG. 2a represents the slope progression of the dose profile over the X axis.

Next in step e) of the method, a center point position of the measuring body 13 is first determined by a maximum point in the course of the dose profile 5 at least one expected limit of the measuring body 13 between a dose maximum 19 and a Dosisminimum 28 a turning point 29 and between a Dosisminimum 28 and a Dosismaximum 20 a turning point 30 is determined. The determination of the inflection points 29 and 30 can be made in the range of the 50% dose point, wherein the 50% dose point is the position in the course of the dose profile 5, at the dose between the Dosisminimum 28 and the Dosismaxima 19 and 20 50%. Preferably, the 50% dose point is determined between two pixels of which a first pixel represents a dose less than 50% and a second pixel adjacent the first pixel represents a dose greater than 50%. Analog can be proceeded to determine the field boundaries of the irradiation field, wherein in the course of the dose profile 5 at least one expected limit of the irradiation field between a dose minimum 18 and a maximum dose 19 a turning point 3 and between a dose maximum 20 and a minimum dose 21 a turning point 4 is determined , The determination of inflection points 3 and 4 may be in the range of 50%

Dose point, wherein the 50% dose point is the position in the course of the dose profile 5, in which the dose between the dose minima 18 and 21 and the Dosismaxima 19 and 20 has 50%. Preferably, the 50% dose point is determined between two pixels of which a first pixel represents a dose less than 50% and a second pixel adjacent the first pixel represents a dose greater than 50%.

In the subsequent step f) of the method, the determined inflection points are assigned to a field boundary or a measuring body boundary. In the example shown, the inflection point 29 of the in negative X-

Direction located body boundary 1 of the measuring body 13 and the inflection point 30 of the lying in the positive X direction body boundary 2 of the measuring body 13 are assigned. Likewise, a determination of the body boundaries of the measuring body 13 in the Y direction can be made on the basis of a dose profile in the Y direction. In FIG. 1, the body boundaries in the Y-direction are identified by the reference symbols 16 and 17. The inflection point 3 can be assigned to the field boundary located in the negative X direction and the inflection point 4 of the field boundary located in the positive X direction. Similarly, a determination of the field boundaries in the Y direction can be based on a dose profile in the Y direction. After determining the two body boundaries of the measuring body 13, the center point position of the measuring body 13 can be determined by the arithmetic mean value formation of the two spatial coordinates of the boundary points. Preferably, the distances of the body boundaries in the X direction and in the Y direction are used for determining the midpoint position. Likewise, after determining the two field boundaries 3 and 4, the central beam position can be determined by the arithmetic mean value formation of the spatial coordinates of the boundary points. Preferably, the distances of the field boundaries in the X direction and in the Y direction are used to determine the central beam position.

In the further step g) of the method, the positions of the center point of the measuring body 13 and of the irradiation field relative to the EPID center are determined on the basis of the associated body boundaries of the measuring body 13 and on the basis of the associated field boundaries of the irradiation field. The steps d) to g) are performed for the directions X and Y. In step h), the difference vector in the EPID plane, which points from the center of the measuring body 13 to the central jet piercing point through the EPID plane, is projected into the isocenter plane according to equation (1). The steps b) to h) are used for all prescribed support arm angle, collimator angle and table angle (des

Patient table).

Finally, the vector components of the difference vector are used to correct the current radiological isocenter.

The method according to the invention achieves a spatial resolution of 0.01 mm with a clinically standard EPID, which corresponds to a 39.2 times higher resolution than the standard EPID.

The measurement conditions for the dose profile shown in Figure 2 were determined as follows: support arm angle = 0 °, collimator angle = 0 °, table angle = 0 °, nominal field size = 15 x 15 mm ² , focus EPID distance = 1.5 m (magnification factor = 1.5 ), Photon energy = 6 MeV and irradiated dose = 12 MU. Dose D is measured by the EPID in the unit [CU] (calibrated dose unit) with 1 CU = 1 Gy under calibration measurement conditions. The sizes a _x and CAX _X are the field width and the central beam position in the X direction, respectively. For the same irradiation field without Winston-Lutz pointer, the same values were determined for these quantities. The graphics were generated using MATLAB ^® (version R2007a).

FIG. 2b shows the definition of all four 50% dose points of the dose profile in the X direction of FIG. 2a, which are required for determining the measuring body boundaries, irradiation field boundaries and the positions of the measuring body center and the central beam. All 50% dose levels are defined as the arithmetic mean of a local dose minimum and a local maximum dose. The 100% dose is the smallest of all local dose maxima. The smallest possible values for the measuring body and for the irradiation field are used as dose minima. If both the respective maximum dose and the respective dose minimum for calculating the 50% dose point is minimal, its dose value also becomes minimal. Every 50% -

Dose point is in the range of the turning point there. Both aspects increase the resolution of the method according to the invention. The fact that a low 50% dose point is an advantage can be explained physically as follows: The lower the 50% dose point of the field, the less disturbing the scattering photons generated in the measuring body, because the distance of the measuring body increases to the 50% dose point. And vice versa, the lower the 50% dose point of the measuring body, the less disturbing the scattering photons generated by the field boundary (slits, slats of an MLC or a circular collimator), since the distance of the field boundary from the 50% dose point increases. FIG. 2b shows a dose profile for the determination of the geometric properties of the irradiation field and the measuring body in the X direction. In the method according to the invention, however, at least two different dose profiles are cut out of a dose image for this purpose, which have different Y coordinates. In the Y direction to proceed accordingly. This also contributes to the good resolving power or to minimizing the error of the method according to the invention. The equations of determination are:

Legend to the equations:

MK = measuring body or Winston-Lutz pointer or tungsten sphere (the Winston-Lutz pointer used in the explanation of an embodiment of the method according to the invention is a commercial one of the company

BRAINLAB AG, Feldkirchen, Germany)

Field = irradiation field with specific field width (X-direction) and field length (Y-direction)

Coordinate system = EPID coordinate system

D _100% = 100% replacement dose (compare 100% dose of standard IEC 60976)

D ₅₀ (field) = 50% dose to determine the field width and field length by means of the localized tangent

D _50% (MK) = 50% replacement dose for determining the measuring body limits by means of the localized turning tangents

D _max (-X) = local maximum dose of X-profile with negative local coordinate D _max (+ X) = local maximum dose of X-profile with positive local coordinate D _max (-Y) = local maximum dose of Y-profile with negative local coordinate D _max (+ Y) = local maximum dose of the Y-profile with positive location coordinate D _min (MK) = local dose minimum of both dose profiles in the area of the measuring body

D _min (-X) = local dose minimum of the X-profile at the edge of the field with negative location coordinate

D _min (+ X) = local dose minimum of the X-profile at the edge of the field with positive location coordinate

D _min (-Y) = local dose minimum of the Y-profile at the edge of the field with negative location coordinate

D _min ( ⁺ Y) = local dose minimum of the Y-profile at the edge of the field with positive location coordinate

D _min (field) = uniform dose minimum of both dose profiles at all field margins

Xi (MK) = boundary of the measuring body in the negative X direction X ₂ (MK) = boundary of the measuring body in the positive X direction

Yi (MK) = boundary of the measuring body in the negative Y direction

Y ₂ (MK) = boundary of the measuring body in the positive Y direction

Xi (field) = position of the field boundary in the negative X direction

X ₂ (field) = position of the field boundary in the positive X direction

Yi (field) = position of the field boundary in the negative Y direction

Y ₂ (field) = position of the field boundary in the positive Y direction

ΔΧ (ΜΚ) = position of the measuring element center in the X-direction relative to the EPID center

ΔΥ (MK) = position of the center of the measuring element in the Y direction relative to the EPID

center

ACAX _X = position of the central beam of the irradiation field in the X-direction relative to the EPID center

ACAX _Y = position of the central beam of the irradiation field in the Y-direction relative to the EPID center

ΔΧ 0 = central beam deviation relative to the measuring element center in the X direction (measured in the EPID plane)

AYiso = central beam deviation relative to the measuring center in the Y direction (measured in the EPID plane)

In order to demonstrate the spatial resolution of 0.01 mm achievable by the method according to the invention, the lamellar positions of an MLC with the smallest possible increment 0.01 mm are varied. The determination of the respective central beam position takes place in a large Wasserpha- ntom MP3 with a high resolution Dosimetriediode E Type 60012 as a dose detector, a two-channel electrometer TANDEM and software ME PHYSTO ^® mc ² from PTW GmbH (Freiburg, Germany). In contrast to the method according to the invention, in which the focus EPID distance is 150 cm, the focus-detector distance for detecting the spatial resolution capability is 100 cm. In carrying out the proof of

Spatial resolution is set to a photon energy = 6 MeV, a dose rate = 400 MU / min, a dose integration time per measurement point = ls and a detector step size = 0.2 mm to 1 mm. In Figure 3, the detection of the spatial resolution is visualized. The radiological determination of the central beam deviations or from the theory

isocenter of a radiotherapy device in the large water phantom represents the gold standard in radiotherapy. The black dots mark the central beam displacements determined by the Winston-Lutz method between a centrally located pair of lamellae (here No. 31) in the X direction. For this purpose, the lamella pairs limiting the irradiation field were shifted in a defined manner; the smallest adjustable increment in the used

MLC "High-Definition 120" is 0.01 mm, and the corresponding regression line has the equation The correlation coefficient according to Pearson is r = 0.9991 at p = 0.0000 (probability of non-correlation) and thus indicates a nearly ideal linear relationship. The measurement results in the large water phantom MP3 are shown as black circles. With both measuring methods, the smallest slat shift of ± 0.01 mm can be resolved. The slightly increasing course of the central beam deviation Y | _S o during the Winston-Lutz test, which lasted 15 minutes here, proves the relaxation of the support arm at 0 ° as a result of a positive bending moment acting on the support arm and the radiation head around the fixed X-axis. The black dots in the Winston-Lutz analysis no longer agree with the black circles of the

In the first case, the medium at the edge of the field and in the field is not homogeneous: air, plastic and tungsten versus homogeneous water. The graph in Figure 3 was generated by MATLAB ^® (version R2007a).

FIG. 4 shows dose profiles of the Winston-Lutz pointer as a measuring body in X-direction 7 and Y-direction 8 with the turning points for defining the measuring body boundaries in X-direction 9 and 10 and in Y-direction 11 and 12. The dashed vertical Lines 7.1 and 8.1 indicate the position of the center of the sphere in the X direction (7.1) and in the Y direction (8.1). The stretch factor is k ^-1 > 1 with the definition in equation (1). The graph in Figure 4 was generated using MATLAB ^® (version R2007a). FIG. 5 shows four dose profiles of a 15 × 15 mm ² MLC-limited irradiation field in the X-direction under the pairs of slats No. 29 to No. 32 with the inflection points 22 and 23 for defining the field boundaries of the irradiation field and with the layers of the local area Central rays (dashed vertical lines 24). The stretch factor is k ^-1 > 1 with the definition in equation (1). The graph in Figure 5 was generated by MATLAB ^® (version R2007a). FIG. 6 shows a further two dose profiles of an MLC-limited irradiation field of the size 15 × 15 mm ² in the Y direction between the closed plate pairs No. 27 and No. 34 with the inflection points 25 and 26 for defining the field boundaries and the positions of the local central rays in FIG Y-direction

(dashed vertical lines 27). The stretch factor is k ^-1 > 1 with the definition in equation (1). The graph in Figure 6 was generated by MATLAB ^® (version R2007a). FIG. 7 shows a schematic illustration of the geometry of a support arm angle-dependent isocentroide of a radiation therapy device Novalis powered by TrueBeam ™ STx (VARIAN Medical Systems, Inc., Palo Alto, CA, USA and BRAINLAB AG, Feldkirchen, Germany). The solid lines in the colors black, dark gray and gray represent the central beam deviations relative to the measuring body in the directions X, Y and Z of the spatially fixed coordinate system. The dashed lines mark the location coordinates of the isocentroids in these directions. The diameters in the axes of the inertial system, their maximum and the location coordinates are also output as numerical values. In addition, the magnitude of the spatial radius vector is shown as a function of the bracket angle (light gray line) and its

Maximum indicated. Analog presentation of results both for the solitary collimator- and romantic angle-dependent Isozentroide and for the global Isozentroid that combines all three solitary Isozentroide, using MATLAB ^® (Version R2007a) generated.

Figure 8 shows a graphical representation of the geometry of the global Isozentroids a medical electron linac Novalis powered by True Beam ™ STx using MATLAB ^®. The solid lines in the colors black, dark gray and gray represent the contact angle dependent centroid beam deviations relative to the Winston-Lutz pointer in the directions X,

Y or Z of the inertial system. The boundaries of the additional negative and positive scattering bands for the superimposed collimator rotation are shown as thin dash-dot lines or dashed lines, respectively. With the addition of the table angle-dependent central beam deviations, the corresponding lines are shown thick. The light gray line shows the angle-dependent curve of the maximum radius of the global len isozentroids. The dotted lines mark the location coordinates of the global isocenter. The diameters in the axes of the inertial system, their maxima and the location coordinates are also output as numerical values. In addition, the spatial distances of the axes of rotation of the collimator and table relative to the support arm rotation axis are shown in the diagram at the bottom right.

Claims

1. A method for EPID-based review, correction and minimization of the isocenter of a radiotherapy device, which at least one rotatable about a table axis patient table, a pivotable about a Tragarmachse support arm, arranged on the support arm emitter head for generating a therapy beam, a rotatable collimator, a Device for projecting a radiological isocenter and a digital acquisition system (EPID) for acquiring dose images by means of the therapy beam, comprising the following steps

b) an irradiation field delimited with the collimator is applied for at least one predetermined angle setting of the support arm, the patient support table and the collimator, and thereby

c) at least one common dose image of the measuring body and of the irradiation field is recorded with the EPID,

d) a dose profile for each direction within an EPID coordinate system is created on the basis of the recorded dose image, and

e) in the course of the dose profile at both expected body boundaries of the measuring body in the X direction of the EIPD coordinate system and at both expected body boundaries of the measuring body in the Y direction of the EPID coordinate system between a local dose minimum and a local maximum dose and between a local maximum dose and a local Dosisminium each determined a turning point and

f) the determined positions of the inflection points are assigned to the body boundaries of the measuring body in the X direction and in the Y direction, g) based on the associated body boundaries of the measuring body in the dose image a position of the center of the measuring body relative to the EPID center is determined, wherein the steps d) to g) are carried out analogously to the field boundaries and the field center of the irradiation field, and

h) a difference vector is determined from a positional deviation of the center point of the measuring body from the EPID center and from a positional deviation of the field center point of the irradiation field from the EPID center, and

i) the vector components of the difference vector are used to correct the current radiological isocentre.

2. The method according to claim 1, characterized in that the method steps b) to h) for the angular degrees of freedom of the patient support table, the support arm and the collimator are performed with an increment of at most 30 °.

3. The method according to any one of the preceding claims 1 and 2,

characterized in that differential vectors are used to determine the size and position of the spatial isocenters from dose images which originate from different angular positions of the patient table, the support arm and the collimator, wherein the vector components of the spatial vectors of the spatial isocenters are used to correct the radiological isocentre ,

4. The method according to any one of the preceding claims 1 to 3, characterized in that the irradiation field is applied taking into account minimum values of a field size, a relaxation time of the support arm, a dose per irradiation field, and / or a focus EPID distance.

5. The method according to any one of the preceding claims 1 to 4, characterized in that the / the inflection point / e in the range of a 50% dose point between a dose minimum and a maximum dose profile and / or in the range of a 50% dose point between a maximum dose and a dose modulus of the dose profile is / are determined.

6. The method according to any one of the preceding claims 1 to 5, characterized in that the inflection point / s in the region of the 50% dose point is preferably set between two pixels

of which a first pixel represents a dose less than 50% and a second pixel adjacent the first pixel represents a dose greater than 50%.

7. Ver n according to any one of the preceding claims 1 to 6, characterized in that the position vectors of the spatial isocenters are used for calibration of a patient positioning system.

8. The method according to any one of the preceding claims 1 to 7, characterized in that the position vectors of the spatial isocenters are used to correct the projection device.

9. The method according to any one of the preceding claims 1 to 8, characterized in that when using a multi-leaf collimator (MLC), the steps d) to h) for each irradiation field limiting lamellae of the MLC are performed.

10. The method according to any one of the preceding claims 1 to 9, characterized in that at a variation of the patient table angle a Tragarmwinkel 0 "is set.

11. The method according to any one of the preceding claims 1 to 10,

characterized in that a global spatial isocenter of the radiotherapy device is determined from the individual central beam deviations in the three spatial directions X, Y and Z of the three spatial isocenters of the support arm, the collimator and the patient support table.

12. The method according to any one of the preceding claims 1 to 11,

characterized in that the method steps d) to i) controlled by a software are performed automatically.

13. The method according to any one of the preceding claims 1 to 12, characterized in that all the spatial isocenters influencing device-specific parameters of the radiotherapy device can be optimized by minimizing a predetermined target function.

14. The method according to any one of the preceding claims 1 to 13,

characterized in that geometric tolerances of the radiotherapy device are quantified and taken into account mathematically.

15. The method according to any one of the preceding claims 1 to 14,

characterized in that the isocenter of the patient support table is alternatively determined at a support arm angle = 0 ° by means of a radiation therapy device independent radiological patient positioning system.

16. Application of the method according to the preceding claims 1 to 15 to a therapy simulator, wherein the spatial isocenters are determined, corrected and minimized.